On November 21, continuing the 10th Session, the National Assembly listened to the report on the Inspection of the draft Law on Artificial Intelligence (AI).
Presenting the review report, Chairman of the National Assembly's Committee on Science, Technology and Environment Nguyen Thanh Hai said that with the view that law-making is a framework law, in general, most of the provisions of the draft Law are consistent with the practical conditions in the early stages of AI development in Vietnam today.
To ensure the feasibility of the Law, the Committee recommends that the drafting agency pay attention to some additional contents, such as continuing to review and improve to ensure that the framework regulations are clear, concise, and have principles and orientations to continue developing guiding documents and effectively applying them in practice.
Complete the synchronization of the draft Decree with detailed regulations and instructions for implementation, other relevant guiding documents and issue a report to the Government for promulgation promptly, avoiding creating legal gaps.
Regarding pre-inspection regulations, the Committee proposed to review and reduce pre-inspection regulations such as requiring technical records and activity logs before putting products into circulation, which will increase compliance costs, slow down innovation and application of AI, reduce competitiveness and attract investment, and need to be studied strongly to switch to the post-inspection mechanism.
Regarding the national database of AI systems and the database for AI, the Committee proposes to add a core principle to ensure the quality of data for AI such as: data must be "correct - sufficient - clean - alive - unified - shared".
There is a mechanism for connecting and sharing data to prevent data from being dispersed, creating bottlenecks in AI research and development; there are mandatory principles on ensuring network security, data security and protecting national AI infrastructure, preventing the risk of taking control and data leakage.
Regarding the legal responsibility of the parties involved, AI can also perform human-made behaviors and errors. Meanwhile, the legal responsibility of AI has many very different views, so it is difficult to determine responsibility in the traditional sense. When an incident arises, it will cause disputes about administrative, civil and criminal responsibilities.
Therefore, it is necessary to supplement the principle of distinguishing responsibilities between entities and stakeholders, including foreign suppliers when providing cross-border AI services, and distinguishing between cases of intentional intention, intentional intention or errors due to technical limitations and unpredictability.
Regarding prohibited acts, according to Chairman Nguyen Thanh Hai, the Committee proposed to add regulations on prohibited acts to contribute to early identification of violations from the research to implementation stage of using AI.
This includes the behavior of users, organizations and individuals who take advantage of AI to carry out prohibited acts in the following direction: Prohibit the use of AI to disrupt, incite politics, violate national security, manipulate votes and elections;
Prohibit the use of AI to create fake content, create images, video clips, serve fraud, insult honor, dignity, cause division and other bad purposes.
